Vol. 2, No. 6, Nov.-Dec., 1963
SYNTHESIS OF HYDROXAMIC ACIDS
1313
TABLEVI1 Dudek, G. O., and Westheimer, F. H. (1959), J . A m . Chem. SOC.81, 2641. HYDROLYSIS OF N-ACETYLDEHYDROALANINE~~~I~ Farrington, J. A., Hextall, P. J., Kenner, G. W., and Turner, k x 104 J. M. (1957), J. Chem. SOC.,1407. PHd (min-1) Folin, O., and Ciocalteu, V. (1927), J. Biol. Chem. 73, 627. 7.4 6.4 5.6 5.0 4.2
0.1 0.5 2.1 5.8 31.6
At 78’ in 25% ethanol-water. * Compound a t 2.7 X M. Rate measured by decrease of absorbance a t 240 mp. p H determined a t 78O. For composition of buffers, see Experimental.
ampoule technique described. Results are given in Table VII. REFERENCES Adler, A. J., Fasman, G. D., and Blout, E. R. (1963),J . A m . Chem. SOC.85, 90. Barnard, P. W. C., Bunton, C. A., Llewellyn, D. R., Oldham, K. G., Silver, B. L., and Vernon, C. A. (1955), Chem. Ind. (London), 760. Ben-Ishai, D. (1956), J . A m . Chem. SOC.78, 4962. Bergmann, M., and Grafe, K., (1930), 2.Physiol. Chem. 187, 187. Bergmann, M., and Zervas, L. (1932), Ber. deut. chem. Ges. 65, 1192. Bernhard, S. A,, Berger, A., Carter, J. H., Katchalski, E., Sela, M., and Shalitin, Y . (1962), J . A m . Chem. SOC.84, 2421. Blumenthal, E., and Herbert, B. M. (1945), Trans. Faraday SOC.41, 611. Bruice, T. C. (1962), Enzyme Models and Enzyme Structure, Brookhaven Symp. Biol. 15, 76. Bruice, T. C., and Benkovic, S. J. (1963), J . A m . Chem. SOC. 85, 1. Cohen, L. A., and Witkop, B. (1961), Angew. Chem. 73, 253. Cohen, T., and Lipowitz, J. (1961), J. A m . Chem. SOC.83, 4866.
Frost, A. A., and Pearson, R. G. (1961a), in Kinetics and Mechanism, New York, Wiley, p. 160. Frost, A. A., and Pearson, R. G. (1961b), in Kinetics and Mechanism, New York, Wiley, p. 166. Fry, E. M. (1949), J. Org. Chem. 14, 887. Fry, E . M. (1950), J. Org. Chem. 15, 802. Greenstein, J. P., and Winitz, M. (1961), in Chemistry of the Amino Acids, New York, Wiley, p. 856. Hejne, H. W. (1957), J. A m . Chem. SOC.79, 907. Iselin, B., Rittel, W., Sieber, P., and Schwyzer, R. (1957), Helv. Chim. Acta 40, 373. Kumamoto, J., and Westheimer, F. H. (1955),J . A m . Chem. SOC.77, 2515. Martin, R. B., and Parcell, A. (1961), J . A m . Chem. SOC.83, 4835. Morawetz, H., and Otaki, P. S. (1962), Unpublished data reported by Bruice (1962), q.v. Patchornik, A., and Sokolovsky, M. (1962), Bull. Res. Council Israel, Sec. A . 11, 80. Peter, H., Brugger, M., Schreiber, J., and Eschenmoser, A. (1963), Helv. Chim. Acta 46, 577. Porter, G. R., Rydon, H. N., and Schofield, J. A. (1960), J. Chem. SOC.,2686. Riley, G., Turnbull, J. H., and Wilson, W. (1957), J . Chem. SOC.,1373. Rothstein, E. (1949), J. Chem. SOC.,1968. Scott, F. L., Gljck, R. E., and Winstein, S. (1957), Experientia 13, 183. Sondheimer, E., and Holley, R. W. (1954), J . A m . Chem. Soc. 76, 2467. Sondheimer, E., and Holley, R . W. (1957),J . A m . Chem. SOC. 79, 3767. Stirling, C . J. M. (1960), J . Chem. SOC.,255. Winstein, S., and Boschan, R. (1950), J . A m . Chem. SOC. 72, 4669. Witkop, B. (1961), Advan. Protein Chem. 16, 265. Woodward, R . B., Olofson, R . A., and Mayer, H. (1961),J. A m . Chem. SOC.83,1010. Zioudrou, C., and Schmir, G. L. (1963); J . A m . Chem. SOC. 85, in press.
Equilibrium Constants for the Synthesis of Hydroxamic Acids”
w. P. JENCKS,M. CAPLOW,? M. GILCHRIST,AND R. G. KALLEN~ From the Graduate Department of Biochemistry, Brandeis University, Waltham, Massachusetts$ Received June 19, 1963 Hydroxamic acid formation from hydroxylamine and unactivated carboxylic acids proceeds readily in aqueous solution a t elevated temperatures a t pH 4 4 or a t 2 5 ” in dilute acid. The equilibrium constants for the formation of acetohydroxamic, hexanohydroxamic, octanohydroxamic, and N-acetyl-L-tyrosine hydroxamic acids have been determined a t 25 ’. At p H 7 the apparent equilibrium constants for the formation of simple hydroxamic acids are near 1, while that for acetyl-L-tyrosine hydroxamic acid is 0.042. The determination of activated acyl groups by their conversion to hydroxamic acids has been an extremely useful analytical tool because of the specificity of
* This work was supported in part by the National Science Foundation, and by a grant (CA-03975) from the National Cancer Institute of the National Institutes of Health, U. S. Public Health Service. t This work was carried out during the tenure of a postdoctoral fellowship awarded by the National Institute of Dental Research. 1 Postdoctoral trainee on a National Institutes of Health training grant (2G-212). B Publication No. 243.
the reaction for different types of acyl groups and the ease with which the ferric complex of hydroxamic acids can be determined spectrophotometrically. The requirements for the formation of a hydroxamic acid from an acyl compound are, fist, that the rate of the reaction be appreciable and, second, that the equilibrium of the reaction be favorable under the particular experimental conditions employed. It is sometimes assumed that hydroxylamine will react only with activated “energy-rich’’ acyl compounds, and that either the rate or the equilibrium constants are unfavorable for reactions of hydroxylamine with “lowenergy” acyl compounds, including carboxylate ions.
1314
Biochemistry
W. P. JENCKS,M. CAPLOW, M. GILCHRIST, AND R. G. KALLEN 5-
0.4
J-
2-
0 d
ln !-
a W
0.3
I-
4
I:
a X
0
a
0
>
I
0.2
0 K a W
L L
/ /
0
P
0
V
z a m a
0
v, m
a
2
I
3
7 9
MINUTES, XlO-'
FIG.1.-Dependence of the rate of acetohydroxamic acid formation on the concentrations of reactants and catalyst a t 25'. Curve 1: 0.25 M HC1,3.46 M CHaCOOH, 1.6 M NHZOH.HC1. Curve 2: Same, 0.5 M HCl. Curve 3: Same, 1.0 M HC1. Curve 4: 0.5 M HC1, 1.73 M CHICOOH, 1.6 M NH,OH.HCl. Curve 5: 0.5 M HC1,3.46 M CHaCOOH, 0.8 M NHzOH.HC1.
At neutral or slightly acidic p H the free base form of hydroxylamine in aqueous solution reacts readily with energy-rich acyl compounds which have a good leaving group, such as acylimidazoles, acyl phosphates, thiol esters, and phenyl esters (Lipmann and Tuttle, 1950; Noda et al., 1953). At alkaline p H hydroxylamine reacts rapidly with esters of aliphatic alcohols in a base-catalyzed reaction (Hestrin, 1949; Chantrenne, 1948; Bergmann and Wurzel, 1953). Esters with a moderately good leaving group or with electronwithdrawing substituents on the acyl portion, such as amino acid esters and the ester group of N,O-diacetylserinamide, react a t a detectable rate with hydroxylamine near neutral or slightly alkaline pH, but do so largely through a base-catalyzed reaction, which can easily be distinguished from the reactions of compounds with a good leaving group, which proceed readily without specific base catalysis even a t p H 5-6 (Bergmann and Wurzel, 1953; Raacke, 1958; Anderson et al., 1961; Anderson, 1962). Amides react on prolonged incubation a t alkaline p H (Bergmann, 1952; Goldenberg and Spoerri, 1958), or, generally a t elevated temperatures, a t slightly acidic p H (Lipmann and Tuttle, 1950; Katz et al., 1953; Meister et al., 1955). The reactions of energy-rich compounds a t neutral p H proceed to a variable extent through the intermediate formation of unstable O-acylhydroxylamines, but no such intermediate has been demonstrated for the other reaction conditions (Jencks, 1958).' Relatively little is known about the equilibrium constants for hydroxamic acid formation. Since hydroxamic acids are synthesized from carboxylate ions and concentrated hydroxylamine in the presence of certain hydrolytic enzymes, including pancreatic and liver esterases, acetylcholine esterase, w-amidase, glutaminase, asparaginase, and an extract from Proteus vulgaris, the equilibrium constant must not
be strongly in favor of hydrolysis a t neutral or slightly acidic pH (Lipmann and Tuttle, 1950; Hestrin, 1950; Wilson et al., 1950; Collier and Solvonuk, 1955; Lynn and Perryman, 1960; Meister et al., 1955; Miller and Waelsch, 1953). Bernhard et al. (1960) have reported a value for the equilibrium constant for hippuryl hydroxamic acid formation. Meister and Wilson2 and their coworkers have recently carried out independent measurements of the equilibrium constants for the formation of y-glutamyl and acetyltyrosine hydroxamic acids and have obtained results in good agreement with those reported in this communication. The equilibrium constant determinations for hydroxamic acid formation which are reported here were carried out as a part of a program to obtain further 1 Although 0-acylhydroxylamines are formed from activated acyl compounds, tne formation of O-acylhydroxylamines from carboxylic acids has not been demonstrated and would be expected to be energetically unfavorable. I n any case, 0-acylhydroxylamines cannot accumulate a t neutral pH in the presence of concentrated hydroxylamine, because they are quantitatively converted to hydroxamic acids under these conditions; there is no rapid interconversion of 0-acylhydroxylamines and hydroxamic acids under the conditions of the ferric chloride assay, as shown by the fact that 0-acylhydroxylamines do not give a color with ferric chloride, even in the presence of hydroxylammonium ion (Jencks, 1958). * Ehrenfeld et al. (1963) have found that the apparent eqiiilibrium constant for the formation of y-glutamyl hydroxamic acid a t p H 7.2 and 37" is 0.24-0.43. This gives values, for the conventions used in this paper, of approximately 350 for K I , 5.6 for KII, 5.6 X lo8 for KIII, and 0.51 for K r p ~ ,These . values are in good agreement with those reported here and show the expected effect of the acidity of the y-carboxyl group of glutamate ( p K , 4.25). Epand and Wilson (1963) have obtained a value of 0.09 for the equilibrium constant for acetyltyrosine hydroxamic acid formation a t p H 6.6. This is in good agreement with the data reported here.
V d . 2,No. 6,Nov.-Dec., 1963
FIG.2.-Pseudo first-orderrate constants for the approach to equilibrium of acetohydroxamicacid formation at 2 5 O as a function of acid concentration. Solid circles, 3.46 M CHICOOH, 1.6 M NHZOH.HC1; open circle, 1.73 M CHZCOOH,1.6 M NHZOH.HC1; triangle, 3.46 M CHICOOH, 0.8 M NHzOH.HC1.
evidence regarding the energy-rich nature, as measured by the free energy of hydrolysis, of different classes of acyl compounds (see Jencks et al., 1960). I n addition, it was of interest to obtain further information regarding the experimental conditions under which hydroxamic acids are formed nonenzymatically from carboxylic acids.
EXPERIMENTAL Hog liver esterase was purified according to Lipmann and Tuttle (1950) through step L-4. The preparation was found to have a specific activity of 5.2 Mmoles .mix~-l.mg-~for the hydrolysis of 0.0036 M ethyl butyrate a t 37O, p H 7.3. a-Chymotrypsin was obtained from the Worthington Biochemical Corporation and was used without further purification. Hydroxylamine hydrochloride was recrystallized from ethanol-water. Hexanoic and octanoic acids were distilled and the concentrations of aqueous solutions were determined by titration. Hexanohydroxamic acid (mp 64 ") and octanohydroxamic acid (mp 78-79 ") were synthesized essentially by the procedure of Safir and Williams (1952) and sodium N-acetyl-Ltyrosine hydroxamic acid, mp 191-191.5 (decomp), was synthesized according to Hogness and Niemann (1953). Acetohydroxamic acid (mp 89-92.5 ") was prepared from methyl acetate and alkaline hydroxylamine in water-methanol a t 0". The alkaline solution was brought to p H 4 with HC1 and evaporated to dryness under reduced pressure. The product was taken up and recrystallized in absolute ethanol. Hydroxamic acid concentration was followed by
SYNTHESIS OF HYDHOXAMIC ACIDS
1315
addition of aliquota of the reaction mixtures to ferric chloride in hydrochloric acid solution and spectrophotoinetric determination of the ferric hydroxamate a t 540 mp (Lipmann and Tuttle, 1945). Hexano- and octanohydroxamic acids were determined by adding the aliquot plus suflicient water to give 2.0 ml to 4.0 ml of a solution containing 10% FeCh.6 HzO, 0.66 M hydrochloric acid, and 3.3% trichloroacetic acid. The samples were centrifuged and read within 2 hours. A small correction (5-15 Klett units) was made for color from the enzyme a¶one. Acetohydroxamic acid was determined by the addition of a 0.5-ml aliquot and 0.5 ml of water to 4.0 ml of 10% FeCL.6 HzO in 0.3 M hydrochloric acid followed by the addition of 1.0 ml of a solution prepared by mixing equal volumes of 4 M hydroxylamine hydrochloride and 3.5 M sodium hydroxide. N-Acetyltyrosine hydroxamic acid formation was measured by the addition of 0 . 2 4 aliquots of the reaction mixture to 4.0 (or 8.0) ml of 20% FeCh.6 HzO in 1.4 M HC1, followed by the addition of 0.1 (or 0.2) ml of 6.1 M trichloroacetic acid, centrifugation, and filtration through glass wool to remove precipitated protein. The concentrations of carboxylic acids and hydroxylamine were obtained from the amounts added to the reaction mixture after correction for any concentration changes due to hydroxamic acid formation or hydrolysis. The fraction of hydroxylamine present as the free base was determined by titration or from the measured p H and the ionization constant, determined a t 25" a t the same ionic strength as the reaction mixture. Determinations of p H were made with a Radiometer PHM 4b p H meter with a G200B glass electrode. The concentration of water in a given reaction mixture was corrected for the volume taken up by other reactants, using the convention that the activity of pure water is 1.0; this correction was small in most experiments. Ethylenediaminetetraacetic acid, 2 X lob4 M, was added in most experiments a t neutral p H to retard heavy metal-catalyzed decomposition of hydroxylamine. The pK, values of the carboxylic acids and hydroxylamine hydrochloride were determined a t 25" by titration, using linear plots of the equation pH = pK' log [base]/ [acid]. Solutions of 0.010 M octanoic acid were made up in excesa base and titrated with acid. Precipitation of octanoic acid occurred after approximately 30% titration so that the pKa' was determined by extrapolation of the data obtained in the initial third of the titration. The pKa' values of hexanoic and octanoic acids were found to be 4.56 and 4.60, respectively, a t ionic strength 1.0, maintained with sodium chloride. The pK,' values of Nacetyltyrosine and hydroxylammonium chloride in 2.0 M sodium chloride were found to be 3.10 and 5.97, respectively, by titration and the pKa' of N-acetyltyrosine hydroxyamic acid in 2.0 M potassium chloride was found to be 8.67, from measurements of the p H values of partially neutralized solutions. The pKa values of other hydroxamic acids are well above the range of p H used in these experiments (Wise and Brandt, 1955). Values of the pK,' of hydroxylamine hydrochloride a t lower salt concentrations were taken from Kurtz and Niemann (1962).
+
RESULTS Acetohydroxamic Acid.-It was shown in preliminary experiments that hydroxamic acids are formed rapidly upon heating aqueous solutions containing hydroxylamine and acetic or formic acids. Meister et al. (1955) have previously demonstrated the nonenzymatic
1316
Biochemistry
W. P. JENCICS, M. CAPLOW, M. GILCHRIm, AND R. G. KALLEN IO
I
R
3.0
5
t
I
.? 2.0 V
a
.-V
E
e
0
r
c 0
1.0
0
c
s
Lo' 0 Hours
FIG. 4.-Approach
5
to equilibrium of octanoic acid, hydroxylamine, and octanohydroxamic acid a t 25.0", catalyzed by liver esterase, 2.0 mg/ml. Total Garboxylate hydroxamic acid = 3 X M; 2 X M EDTA. Solid circles, p H 7.24 f 0.02, 0.97 M hydroxylamine; open circles,pH 7.22 i0.02,0.49 M hydroxylamine.
+
IO
MINUTES,XIO-~
FIG.3.-Approach to equilibrium for acetohydroxamic acid formation in 0.25 M HC1 a t 25'. Curves 1-3: 3.46 M CH&OOH, 1.60 M NH?OH.HCl, CHJCONHOH as indi1.0 M KC1. Curve 5: Same as cated. Curve 4: Same, 1-3 except 0.80 M NH,OH.HCl. Curve 6: Same, except 1.73 M CHsCOOH.
+
formation of hydroxamic acids from dicarboxylic acids in aqueous solution a t elevated temperature. The rate of hydroxamic acid formation from acetic acid was found to show a maximum a t p H 5.0, which suggests that free acetic acid and hydroxylamine are the reactive species. The rate with formic acid was found to be five to ten times faster than that with acetic acid. It was shown that the equilibrium constant for acetohydroxamic acid formation was larger than 15 (based on the concentrations of uncharged reactants and products), but it was not possible to obtain an accurate value for the equilibrium constant because it was not possible to approach equilibrium from both directions. Experiments were therefore carried out a t 25" with the help of enzymatic or acid catalysis to approach equilibrium. Aqueous acetic acid and hydroxylammonium chloride react spontaneously in the presence of a catalytic amount of hydrochloric acid to form acetohydroxyamic acid (Fig. 1). The reactions follow pseudo first-order
kinetics for the approach to equilibrium. Both the rate and the extent of hydroxamic acid formation are increased by increasing the acetic acid or hydroxylamine hydrochloride concentrations. Increasing the concentration of hydrochloric acid causes a decrease in the equilibrium concentration of acetohydroxamic acid, but also decreases the time required to reach equilibrium. The pseudo first-order rate constants increase linearly with acid concentration (Fig. 2), although this increase does not reflect an actual increase in the rate (in moles per minute) of hydroxamic acid formation (Fig. 1). Similarly, the pseudo firstorder rate constants do not depend on the acetic acid or hydroxylamine hydrochloride concentrations (Fig. 2) although the actual rates of hydroxamic acid formation increase with increasing concentration of the reactants (Fig. 1). This behavior results from the fact that the reactions are going to equilibrium, rather than to completion, 60 that the observed rates reflect the reaction rate constants in both directions, k1 and k-1, rather than the forward rate constant k1, alone. For a first-order reactiort, the observed rate constant, Kobe, is equal t o k1 Ll. The situation is more complicated here because the reaction in the forward direction is second order, but since a t equilibrium the
+
TABLEI EQUILIBRIUM CONSTANTS FOR ACETOHYDROXAMIC ACIDFORMATION AT 25" IN 0.25 0 CHsJOOH Runa 1 2 3 4d 5 6
M
3.46 3.46 3.46 3.46 3.46 1.73
Equilibrium Concentrations NHzOH .HCI M
1.61 1.61 1.60 1.60 0.80 1.60
M HCl
0
CH3ANHOH Y x 103
Fraction Hz0
x 104
5.40 0.76 5.30 0.76 5.11 0.76 4.13 0.76 2.90 0.79 2.80 0.86 Average (excluding 4)
See Fig. 3. * K A = [CH&ONHOH][H+l fr. H20/[NHrOH+][CH3COOH]. Kr [",OH 1 [CHICOOHI. [",OH 1 estimated from measured p H and p K a ' ",OH + = 6.00.
1.84 1.81 1.75 1.42 2.07 2.18 1.93 =
Krc 425 426 410 460 360 390 402
[CHICONHOH] fr. H20/. 1.0 M KCl added.
Vol. 2,No. 6,Nov.-Dec., 1963 EQUILIBRIUM CONSTANTS FOR HEXANOIC AND _
_
_
___ ~
~
~
TABLE I1 HYDROUMIC ACID FORMATION AT 25'
OCTANOIC
C8
cs
IONICSTRENGTH 1.0
~
NHtOHa
C8 CS CS Ce
AND
~
Equilibrium Concentrations Acid C8 CS
PH
RCOOHa
RCONHOH
(M x
(M X loa)
10') 1.70 2.10 1.30 1.82 7.00 8.16 3.06 4.02
(MI
7.24 7.21 7.04 7.03 7.56 7.50 7.19 7.18
0.97 0.49 1.00 0.50 0.98 0.49 1.00 0.50
1.30 0.90 1.70 1.18 3.00 1.84 2.94 1.98
Refers to total concentration of all ionic forms Dresent in solution. uncharged reactants and products.
Time to Obtain Equilibrium (hours) 14.5 14.5 10.0 10.0 10.0 10.0 21.5 21.5
~ _ _ _ _ _ _ _
Ktb 0.76 0.85 1.26 1.28 0.42 0.45 0.93 0.97
N
5
X
1 1
.-w 4
.-0
E
O
w 2 s I
3
0)
.-c
--
; 2 s
-a 0 0
4
1
P
I
I
I
I
I
I
5
IO
IS
20
25
30
I
Hours
FIG.5.-Approach
352 377 380 378 433 404 420 430
Experimental value, at the indicated pH.
6
2
KI'
c
For
Hexano- and Octunohydroxamic Acids. -The approach to equilibrium for octanohydroxamic acid at p H 7.03 and 7.22, catalyzed by liver esterase, is shown in Figure 4. The same equilibrium position is approached from both directions and is maintained in a reaction mixture in which the initial concentrations of reactants were near the expected equilibrium values. These results and the results of similar experiments with hexanohydroxamic acid are summarized in Table 11. These equilibrium constants are given in terms of K', which is the apparent equilibrium constant a t a given p H value and refers to the total of all ionic species of each reactant, and in terms of KI, which refers to only the uncharged species of the reactants and products and is independent of p H . As shown in Table 11, the values of K ' do depend on the p H as expected, while those of KI are independent of p H , within experimentalerror, in the range of p H examined. N-Acetyl-L-Tyrosine Hydroxamic Acid. -The approach to equilibrium of the chymotrypsin-catalyzed reaction of N-acetyl-L-tyrosine and hydroxylamine to form the corresponding hydroxamic acid is shown in Figure 5 . The same equilibrium position is approached from both directions a t each of two p H
reaction proceeds only a very short distance toward completion, the observed rates will be dominated by the iirst-order rate constant for the reverse reaction k- 1. The approach to equilibrium from two directions is shown in Figure 3. The equilibrium concentration of hydroxamic acid is decreased by the addition of 1.0 M potassium chloride. The results are summarized in Table I. The equilibrium constants are given in terms of Ka, based on the concentrations of the acidic species of the reactants, and in terms of KI, based on the concentrations of the uncharged species of the reactants. These equilibrium constants do not provide an exact measure of the expected tendency for acetohydroxamic acid formation in dilute aqueous solution, because of the appreciable deviations from ideality in these solutions, which are demonstrated by the effects of variations in potassium chloride, hydroxylamine hydrochloride, and acetic acid concentrations on the equilibrium constants. The data suggest, however, that these activity coefficient effects are not so large as to invalidate the results for most purposes, and this conclusion is supported by the similarity of the equilibrium constants (KI) for aceto-, hexano-, and octanohydroxamicacid formation (Tables I and 11).
2
1317
f3YNTHESIS OF HYbROXAWC ACIDS
to equilibrium of acetyl-L-tyrosine, hydroxylamine, and acetyl-L-tyrosine hydroxamic acid a t 25O, catalyzed by chymotrypsin. The numbers refer to Table 111, in which the experimental conditions are given.
1318
Biochemistry
W. P. JENCKS, M. CAPLOW, M. GILCHRIST,AND R. 0 . W E N
EQUILIBRIUM CONSTANT FOR
AT
Reaction No. 1 2 3 4 5 6
TABLEI11 SYNTHESIS OF N-ACETYLTYROSINE HYDROXAMIC ACID 25' AND IONIC STRENGTH 2.0'
THE
Equilibrium Concentrations RCOOHb RCONHOH
"*OH' (M)
(M)
(M)
Final PH
K'C
1.11 1.10 1.10 0.538 0.529 0.533
0,146 0,141 0.159 0.285 0.281 0.289
0,0147 0.0155 0.0140 0.0275 0.0310 0.0268
6.67 6.59 6.67 6.22 6.14 6.21
0.087 0.096 0.077 0.177 0.204 0.171
K I ~ 388 366 342 363 373 348 363
Average: Reaction catalyzed by 2.4 x 10-4 M chymotrypsin. Ionic strength maintained with NaC1. Sum of all ionic species in solution at equilibrium. c Experimental value, at the indicated pH. For uncharged reactants and products. TABLEIV EQUILIBRIUM CONSTANTSFOR HYDROXAWC ACIDFORMATION AT 25' EXPRESSED ACCORDING TO DIFFERENT CONVENTIONS
K1 K1l = K1ll =
K'p~7=
[RCONHOH][HzO] [RCOOH1..["?OH 1. same [RCONHOH][HzO] [RCOO-1 [",OH +I [RCONHOH][HzOI~ R C O O ~ ~ [ N H]z[H O +H] [RCONHOH][HzO] [RCOO-1 [",OH],,
Water Activity Taken as
Acetohydroxamic
Hexanohydroxamic
Octanohydroxamic
N-Acetyltyrosine Hydroxamic
1.0
402
422
372
363
55.5
22,300
23,400
20,600
20,200
1.0
16
~
1.0
1.60
1.0
values, with minor variations due largely to differences in the pH values of the different reaction mixtures. The results are summarized in Table I11 according to the conventions described.
DISCUSSION The equilibrium constants for the formation of the hydroxamic acids examined in this study are summarized in Table IV. They are expressed according to several different conventions, each of which is useful for certain purposes. All the values are based on concentrations, rather than activities; i.e., they are not thermodynamic equilibrium constants. The equilibrium constants for hydroxamic acid synthesis reported here are several orders of magnitude smaller than those reported by Bernhard et al. (1960) for the synthesis of hippuryl hydroxamic acid. However, it is not certain from the data presented by these workers that equilibrium was reached in their experiments, and the possibility might be considered that their data reflect the relative rates of chymotrypsin-catalyzed hydrolysis and hydroxylaminolysis of methyl hippurate rather than a true equilibrium. The equilibrium constant expressed according to convention (I), which utilizes the concentrations of nonionized reactants and prodncts, is useful for theoretical comparison of the affinities of different groups for each other. Carpenter (1960), in particular, has RCOOH + H2NR' e RCONHR' + HjO [RCONHR'1 [HJOI K 1 = [RCOOH][H?NRT (I) advocated the use of a convention of this kind as a standard. In the hydroxamic acid series, it is of interest that the equilibrium constants for hydroxamic
15.4
x 107 1.46
1.54
x
107
1.40
14.7 1.47
x
107
1.34
0.49 4 . 6 X 1oj 0.042
acid formation from simple fatty acids and from acetyltyrosine are the same according to this convention, although the amounts of hydroxamic acid in equilibrium with a given amount of hydroxylamine and carboxylic acid are quite different for the different acids under most experimental conditions. N-Acetyltyrosine is some 30-fold stronger an acid than the fatty acids, reflecting the electron-withdrawing effect of the acylamino and phenyl groups. The fact that the equilibrium constants for hydroxamic acid formation from these different types of acids are very similar, i.e., that the equilibrium constant is not changed by electron-withdrawing groups, means that the groups -COOH and -CONHOH must have a similar sensitivity to the effects of polar substituents. For example, if the charge distribution in -COOH and -CONHOH were such that carboxylic acids are more destabilized by electron-withdrawing substituents than are hydroxamic acids, one would expect the equilibrium constant for acetyltyrosine hydroxamic acid formation to be larger than that for acetohydroxamic acid formation. Since the carbonyl group is considerably more electronegative than hydrogen and the resonance
-0 I + structure RC=NR,' is important in amides, it would be expected that reaction (I) would be favored by electron-donating substituents on the amine group. The equilibrium constants for the synthesis of glutamine, propionamide, N-methylpropionamide, N-dimethylpropionamide, and the peptide bonds of benzoyltyrosylglycylamide and benzoyltyrosylglycylanilide are 272, 1,310, 1.5 X lo6, 3.0 X 104, 6,600, and 6,000, respectively, calculated from the data of Benzinger et al. (1959), Morawetz and Otaki (1963),
Vol. 2, No. 6, Nov.-Dec., 1963 Dobry et al. (1952), and Gawron et al. (1961).3 On the other hand, preliminary data obtained by Higuchi and Miki (1961) suggest that the equilibrium constant for anilide formation from polycarboxylic acids is near 1.0.4 While the relative magnitudes of several of these values cannot be easily accounted for, there is a tendency for the equilibrium constants to increase as the substituents on the amide nitrogen atom become more electron-donating. It is, therefore, of interest that hydroxamic acids, derived from hydroxylamine, fall in the same class as glutamine and propionamide, which are derived from ammonia, in spite of the strong electron-withdrawing properties of the hydroxyl group. This may be another reflection of the fact that hydroxylamine has a greater tendency than most other amines to add to the carbonyl group (Jencks, 1959; Jencks and Carriuolo, 1960). An estimate of the relative affinity of the carbonyl group for amines and for oxygen-containing compounds may be made if the equilibrium constants are expressed in terms of the actual molar concentration of water, 55.5, rather than according to the convention that water activity is 1.0. This raises the equilibrium constants for hydroxamic acid formation to approximately 20,000 and those for the amides of several more basic amines to even higher values. This means that, on a molar basis, the carbonyl group prefers to form a bond with hydroxylamine over a bond with water by a factor of approximately 20,000, when it is presented with both molecules in aqueous solution. The corresponding value for carboxylic esters is near one, indicating that the carbonyl group has an approximately equal affinity for water and for an alcohol, as would be expected from the chemical similarity of water and alcohols. This difference presumably reflects the low electronegativity of nitrogen and the large ability of the nitrogen atom to donate electrons by resonance. In spite of its usefulness for some puposes, we do not feel that convention I should be used as a general standard for the following reasons: (a) For many reactions, convention I describes a situation that can never be realized experimentally, because there is no pH value a t which all the reactants and products of a reaction are in the nonionized form. For example, a nonionized carboxylic acid cannot exist in the presence of most nonionized amines; one or both will be ionized a t any attainable pH value. (b) In order t o determine the value of KI for a given reaction from experimental measurements, it is necessary to know the dissociation constants of all groups in the reactants and products which are appreciably ionized under the I n calculating the equilibrium constant for glutamine formation according to this convention, only those groups which are directly involved in the reaction, Le., ammonia and the y-carboxyl group of glutamate, were corrected to the uncharged form. The calculations for propionic acid derivatives are based on a p K , ‘ value of 4.6 for propionic acid. The calculations for N-methylpropionamide are based on a corrected value of 4.4 kcal,/mole for AF& (H. Morawetz, personal communication; see Morawetz & Otaki, 1963). The formation of these anilides was reported to take place readily only with polycarboxylic acids, suggesting that the reaction is facilitated by a carboxyl group other than that which is making a new bond with the amine. The fact that hydroxamic acid formation takes place readily with hydroxylamine and monocarboxylic acids indicates that such facilitation is not necessary for amide formation. A similar conclusion may be reached from the experiments of Morawetz and Otaki (1963) on formation of simple amides from monocarboxylic acids, although these reactions are considerably slower.
SYNTHESIS OF HYDROXAMXC ACIDS
1319
experimental conditions chosen for study, in order that the equilibrium concentrations of the nonionized species may be used to calculate KI. These dissociation constants are often unknown or very difficult to determine, particularly in the case of compounds which exist as zwitterions. (c) Convention I becomes unwieldy when it is necessary to deal with molecules containing several ionizable groups, not all of which are directly involved in the reaction under consideration. Either all groups on a molecule must be converted to the nonionized form, which is very difficult for molecules containing both acidic and basic groups, such as ATP, coenzyme A, or amino acids; or only the reacting groups are converted to the nonionized form, in which case the molecules are not completely uncharged. Convention I1 is a particularly convenient pHindependent convention for expressing the equilibrium constant for amide synthesis, because the ionic species
+ HINR’ e RCONHR’ + H 2 0 I
RCOO-
=
[RCONHR’] [HzO] [RCOO- ] [HsN+R’]
(I1)
given are those which actually exist in solution over a considerable range of p H in the p H region often used in experimental work. For hydroxamic acid synthesis it does not correspond directly to an attainable experimental situation, because hydroxylamine is not completely protonated a t any pH value a t which the carboxyl group is fuUy ionized, but i t is still useful for comparative purposes. According to this convention, electron-withdrawing s.ubstituents on the acyl group will destabilize the products more than the starting carboxylate ion and will therefore drive the equilibrium position to the left. (We regard this as a more correct way of expressing the situation than to ascribe the same facts to the greater acidity of acids containing electron-withdrawing substituents. The latter procedure implies a particular reaction path, which is not desirable in a thermodynamic comparison.) The expected effect is seen in a comparison of the equilibrium constants, K I I (Table IV), for hydroxamic acid synthesis from simple fatty acids (approximately 15) with that for acetyltyrosine hydroxamic acid synthesis (0.49). The electron-withdrawing effect of the side chain in acetyltyrosine is reflected in the 30-fold greater ionization constant of this acid compared to acetic acid. Electron-withdrawing substituents on the amine will tend to drive equilibrium I1 to the right, because such substituents will destabilize the positively charged ammonium ion more than the amide. The expected effect is seen in the equilibrium constants for simple as amides (0.028-1.4) and glutamine (3.1 X well as the peptide bonds of benzoyltyrosylglycinamide (0.5) and benzoyltyrosylglycylanilide (O.l), which are smaller than those for the synthesis of simple hydroxamic acids. This convention most clearly expresses the previously mentioned experimental fact that the synthesis of hydroxamic acids is considerably easier a t neutral or slightly acid p H than is the synthesis of most amides. Convention I11 is given because it is the easiest for calculating values of Kfappor AFO’ for hydroxamic acid formation a t different pH values a t neutral or slightly basic pH. In the absence of appreciable ionization of hydroxylamine or hydroxamic acid and under RCOO-
+ HzNR’ + H + e RCONHR’ + Hz0
conditions in which the carboxylic acid is completely
1320
w.
ionized it is only necessary to substitute the hydrogenion concentration to obtain these constants. Convention IV, the equilibrium constant based on the total stoichiometric concentrations of all ionic acid
Biochemistry
P. JENCECS, M. CAPLOW, M. GILCHRIST, AND R. G . KALLEN
+ amine
amide
+ HzO
species of the reactants and products a t a given p H value, is the simplest convention to deal with experimentally, because it does not require knowledge of the dissociation constants of the reactants and products. It is this constant, K', that is generally obtained from experimental measurements. This apparent equilibrium constant will vary as the p H is changed, because it does not take into account the ionization of the reactants and the uptake or release of hydrogen ions. Nevertheless, this convention is very useful for comparisons a t a particular p H value, because it gives directly the ratio of the concentrations of reactants and products a t that pH value. Comparisons a t p H 7.0 are particularly useful, and the values of the free energy of hydrolysis a t pH 7.0, AF"'p~.r,are often used to compare the energy-rich nature of acyl and phosphoryl compounds. The value of K'pH7 near one for fatty acid hydroxamic acids expresses the surprising fact that in the presence of 1 M hydroxylamine a t neutrality a carboxylic acid exists approximately one-half as the hydroxamic acid and one-half as the carboxylate ion a t equilibrium; i.e., the free energy of hydrolysis (AFO') of hydroxamic acids a t p H 7 is near zero (actually f200 calories/mole). For acetyltyrosine hydroxamic acid, KIsPpa t p H 7 is 0.042 and AFo' (hydrolysis) is -1870 calories/mole, reflecting the effect of the electron-withdrawing groups on the acyl moiety of this compound. Although there is only a small thermodynamic barrier to the formation of hydroxamic acids from even the least activated of all acyl g~oups,the carboxylate ion, there remains a kinetic barrier. It is by making use of this kinetic barrier by careful regulation of the reaction conditions, as indicated in the introduction, that the formation of hydroxamic acids from different classes of activated acyl groups may be used as a specific analytical method for the analysis of such groups. ACKNOWLEDGMENTS We are grateful to Miss Gloria Deutsch, Mr. Charles Allen, and Mr. Ray Adman for carrying out preliminary experiments. We wish to thank the National Science Foundation and the National Cancer Institute of the National Institutes of Health for financial support.
REFERENCE! Anderson, B. M. (1962),Biochemisby 1, 1107. Anderson, B. M., Cordes, E. H., and Jencks, W. P. (1961), J . Biol.Chem. 236,455. Benzinger, T.,Kitzinger, C., Hems, R., and Burton, K. (1959),Biochem. J . 71,400. Bergmann, F. (1952),A d . Chem. 24, 1367. Bermann. F.. and Wurzel. M. (1953). . .. Biochim. B i o-~ h- v s . A;ta 12,.412. Bernhard. S. A.. Coles. W. C.. and Nowell. J. F. (1960). . J . Am.'Chem. Soc. 82,3043. ' Carpenter, F. H. (1960),J . A m . Chem. SOC.82,1111. Chantrenne, H. (1948),Compt. Rend. Trav. Lab. Carlsberg 26,231. ( C A , 42,6393). Collier, H. B., and Solvonuk, P. F. (1955),Biochim. Biophys. Acta 16, 583. Dobry, A., Fruton, J. S., and Sturtevant, J. M. (1952),J. Biol. Chem. 195,149. Ehrenfeld, E., Marble, S. J., and Meister, A. (1963), J. B w l . Chem. 238, 3711. Epand, R. M., and Wilson, I. B. (1963),J . Biol. Chem. 238. PC3138. Gawron, O., Glaid, A. J., 111, Boyle, R. E . , and Odstrchel, G. (1961),Arch. Biochem. Biophys. 95,203. Goldenberg, V., and Spoerri, P. E . (1958),Anal. Chem. 30, 1327. Hestrin, S. (1949),J. Biol. Chem. 180,249. Hestrin, S. (1950),Biochim. Biophys. Acta 4,310. Higuchi, T., and Miki, T. (1961),J . A m . Chem. SOC.83, 3899. Hogness, D . S.,and Niemann, C. (1953),J . A m . Chem. SOC. 75,884. Jencks, W. P. (1958),J . A m . Chem. SOC.80,4581, 4585. Jencks, W. P. (1959),J . A m . Chem. SOC.81,475. Jencks, W. P., and Carriuolo, J. (1960),J . A m . Chem. Soc. 82,1778. Jencks, W. P . , Cordes, S., and Carriuolo, J. (1960),J . Biol. Chem. 235, 3608. Katz, J., Lieberman, I., and Barker, H. A. (1953),J . Biol. Chem. 200, 417. Kurtz, A. N., and Niemann, C. (1962),Biochemistry I, 238. Lipmann, F., and Tuttle, L. C. (1945),J . Biol. Chem. 159, 21. Lipmann, F.,and Tuttle, L. C . (19501,Biochim. Biophys. Acta 4 , 301. Lynn, W. S., Jr., and Perryman, N . C. (1960), J . Biol. Chem. 235, 1912. Meister, A., Levintow, L., Greenfield, R. E . , and Abendschein, P.A. (19551,J . Biol. Chem. 215,441. Miller, A., and Waelsch, H . (19531,Federation Proc. 12,246. Morawetz, H., and Otaki, P.S. (1963),J . A m . Chem. SOC. 85,463. Noda, L. H., Kuby, S. A., and Lardy, H. A. (1953),J . A m . Chem. SOC.75, 913. Raacke, I. D . (1958),Biochim. Biophys. Acta 27,416. Ramachandran, L. K., and Narita, K. (1958),Biochim. Biophys. Acta 30,616. Safir, S. R., and Williams, J. H. (1952),J . Org. Chem. 17, 1298. Wilson, I. B . , Bergmann, F., and Nachmansohn, D . (1950), J . Biol. Chem. 186,781. Wise, W. M.,and Brandt, W. W. (1955),J . A m . Chem. SOC. 77, 1058. I.
